Limited recruitment during relaxation events: Larval advection and behavior in an upwelling system
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Limnol. Oceanogr., 57(2), 2012, 457–470
E 2012, by the Association for the Sciences of Limnology and Oceanography, Inc.
doi:10.4319/lo.2012.57.2.0457
Limited recruitment during relaxation events: Larval advection and behavior in an
upwelling system
Steven G. Morgan,a,b,* Jennifer L. Fisher,a Skyli T. McAfee,a,1 John L. Largier,a,b and Chris M. Halle a
a Bodega Marine Laboratory, University of California Davis, Bodega Bay, California
b Department of Environmental Science and Policy, University of California Davis, Davis, California
Abstract
We capitalized on a long-term record of larval recruitment and a distinctive oceanographic signature to reveal
how changes in ocean conditions affect larval advection, behavior, and recruitment in a region of strong, persistent
upwelling and recruitment limitation. We repeatedly sampled the vertical and horizontal distribution of a larval
assemblage and ocean conditions during infrequent relaxations of prevailing upwelling winds near Bodega Bay,
California. During prolonged relaxation events, a poleward, coastal, boundary current imported low-salinity
surface waters that were devoid of larvae to the study area. The resident larval assemblage was restricted to cold,
saline, bottom waters and pushed offshore while diel vertical migrations were suppressed. Hence, changes in
oceanographic conditions strongly affected behaviorally mediated larval distributions, revealing the reason that few
species recruit during relaxation events in this region. During relaxation events in upwelling regions, poleward
coastal boundary currents will force larvae offshore throughout the water column at small headlands and in
seaward-flowing bottom currents at straight coastlines, but recruitment may decrease markedly only near estuaries
where few larvae will arrive in low-salinity surface waters. Targeted profiling of larval assemblages complements
widespread monitoring of recruitment from shore and is necessary to determine how changing ocean conditions
affect larval distributions, recruitment dynamics, and the connectivity of populations.
Complex life cycles in a dynamic ocean make it especially or expedite transport from natal populations (Queiroga and
challenging to establish processes that regulate marine Blanton 2005; Morgan 2006; Morgan and Anastasia 2008).
populations and communities (Morgan 2001; Strathmann The extent of cross-shelf migrations and the timing of
et al. 2002). The sheer numbers of larvae produced, their recruitment onshore largely are determined by the vertical
poor swimming capabilities and the episodic nature of distributions of larvae in a stratified water column (Queiroga
settlement events have led to the widespread belief that high and Blanton 2005; Morgan et al. 2009a,c). However, these
larval mortality results in unpredictable recruitment in time surveys cannot resolve alongshore transport, because larvae
and space (Thorson 1950; Caley et al. 1996). This view was could have originated from many possible locations.
based on logical inference and data from experimental A long-term recruitment study by our colleagues and us
designs that were limited in their ability to determine the exemplifies the reason that complementary profiling of
fate of long-lived microscopic larvae in ocean currents. larvae and ocean conditions across oceanographic events is
Tracking large fluctuations in the numbers of recruits at a necessary to determine the extent of larval advection and
focal study area cannot resolve whether most larvae died or control in regulating alongshore transport and recruitment
were transported to neighboring locales beyond this area of to coastal populations and communities. Although larval
inference (Morgan 1995). In contrast, sampling over large recruitment has been monitored for 13 yr in a region of
spatial and long temporal scales has demonstrated that strong, persistent upwelling on the west coast of the USA,
larval production and recruitment can be coupled (Ro- processes responsible for inter-annual variation in recruit-
bertson et al. 1988; Hughes et al. 2000) and oceanographic ment to coastal populations and communities are still not
patterns can result in predictable spatial changes in resolved (Wing et al. 2003; Mace and Morgan 2006a;
populations and communities (Yoshioka 1982; Broitman Morgan et al. 2009a). Prevailing northwesterly winds force
et al. 2008; Shanks et al. 2010). surface waters offshore due to the Coriolis effect (Ekman
Monitoring larval recruitment at appropriate scales is transport), lowering sea level at the coast and forcing
most feasible from shore for intertidal and shallow subtidal upwelling of bottom waters and southward alongshore
communities, but surveying larval distributions at sea is transport (Largier et al. 1993). Larvae in surface waters
needed to provide a deeper understanding of the processes may be passively advected southward and offshore except
behind the observed patterns. A long history of surveying in the lee of headlands, where eddies may entrain them
ontogenetic changes in the cross-shelf abundances of larvae during prevailing upwelling conditions (Wing et al. 1998;
of nearshore species has shown that exploitation of Lagos et al. 2005; Morgan et al. 2011). When winds
circulation patterns enables larvae to either remain nearby infrequently weaken (relax) or reverse, larvae that have
accumulated in the lee of a large headland (Point Reyes)
* Corresponding author: sgmorgan@ucdavis.edu during upwelling conditions are transported poleward by a
coastal boundary current, which is both characterized and
1 Present addresses: California Ocean Science Trust, Oakland, enhanced by warm, low-salinity surface waters originating
California in the large estuary of San Francisco Bay, which is located
457458 Morgan et al.
, 50 km to the south (Largier et al. 1993, 2006; Wing et al.
2003). Only larvae of the Dungeness crab (Cancer magister)
appear to be transported poleward long distances and
primarily when relaxation events lasting several days occur
during the recruitment season in this region (Wing et al.
2003; Morgan et al. 2009a).
The objective of our present study was to provide a fuller
mechanistic understanding of the dynamic interplay
between larval behavior and ocean conditions on recruit-
ment over upwelling-relaxation cycles. We tested the
hypothesis that larvae are transported poleward by the
coastal boundary current from the larval retention zone in
the lee of Point Reyes during relaxation events, increasing
larval supply to recruitment-limited populations (Wing
et al. 2003). If so, then larvae should be more abundant in
poleward-flowing, warm, low-salinity, surface waters dur-
ing relaxation events than in cold, saline waters during
prevailing upwelling conditions. Larvae must remain within
the surface boundary current to be transported poleward,
but it is unknown whether larvae residing beneath the
surface intrusion will ascend into the warm, low-salinity
waters, especially during diel vertical migrations to the
surface at night (Morgan and Fisher 2010). To test these
hypotheses, we repeatedly profiled the vertical and
horizontal distributions of nearly the entire larval assem-
blage of nearshore benthic crustaceans relative to ocean Fig. 1. Locations of three replicate transects sampled in
conditions for 2 months during the peak upwelling season, Bodega Bay, California, USA, and three replicate transects
over alternating upwelling and relaxation conditions and sampled in the open coastal waters flowing past Bodega Head.
diel cycles. We sampled a small embayment where larvae Sampling occurred during 7 weeks from 07 June to 10 August
are entrained during upwelling (Roughan et al. 2005; Mace 2005. Solid squares depict the locations of the acoustic Doppler
current profilers (ADCP) located in Bodega Bay and off Bodega
and Morgan 2006b; Morgan et al. 2011), as well as in the
Head. Salinity was measured continuously at the Bodega Marine
open waters of the adjacent continental shelf, to determine Laboratory (BML).
the effect of the boundary current on larval distributions
throughout nearshore coastal waters along topographically
complex coastlines in upwelling regions. the mouth of each net to determine the volume of water
sampled.
Methods A concurrent study determined whether larval densities
increased in Bodega Bay upon the arrival of the coastal
Sampling—The study was conducted in the windiest boundary current. We sampled along the three bay
region on the west coast of North America, where transects on alternate days from 07 June to 10 August
equatorward winds persist during the spring and summer 2005, resulting in a time-series of 33 sets of replicate
(Dorman et al. 2006). To determine whether the vertical samples (Fig. 1). We collected larvae throughout the water
and horizontal distributions of larvae changed across the column (10–15 m) with a sled-mounted 0.5-m-diameter ring
study area over upwelling-relaxation cycles, we surveyed net fitted with 335-mm mesh and a flow meter.
the assemblage of crustacean larvae in Bodega Bay (in the During all cruises, profiles of water-column properties
lee of Bodega Head) and along the open coast about 1 km were taken at the beginning and end of each plankton tow to
offshore of Bodega Head from 07 June to 05 August 2005. place larval distributions in the context of hydrographic
Eight surveys were conducted along three replicate patterns. Temperature, salinity, fluorescence, and turbidity data
transects inside the bay and three more replicate transects were collected using a profiling conductivity–temperature–
outside the bay (Fig. 1). Sampling was conducted during depth (CTD) package (Seabird Electronics SBE-19 Plus
the daytime, and it also was conducted at night on two and WETLabs WETStar fluorometer, and WETLabs
consecutive dates (04–05 Aug). Larvae were collected using transmissometer).
an electronically tripped Tucker Trawl (0.5 m2 mouth) that Salinity was measured continuously at the seawater
was equipped with 335-mm mesh nets and real-time intake of the Bodega Marine Laboratory (Seabird Elec-
temperature and depth sensors that allowed samples to be tronics SBE-16 Plus). In addition, moorings obtained time-
collected relative to the thermocline. The water column was series of temperature and current velocity: one mooring
partitioned into two depth bins that were sampled equally. was located at a depth of 10 m in Bodega Bay and a second
In addition, a neuston net (0.5-m2 mouth) sampled the mooring was located in 30 m of depth offshore of Bodega
surface of the water column. Each net was towed for 5 min. Head (Fig. 1). Temperature (T, 6 0.2uC) was recorded
A flowmeter (General Oceanics model 2030) was fitted to every 6 min using Optic StowAwayH TidbiT thermistors.Upwelling and larval recruitment 459
The thermistors were attached to the mooring 1 m above using PRIMER software (version 6.1.10; Plymouth Rou-
the bottom and 1 m below the surface in Bodega Bay and tines in Multivariate Ecological Research). For these
1 m, 9 m, 16 m, 23 m, and 30 m above the bottom offshore analyses, temperature and salinity from CTD profiles and
of Bodega Head. Decreasing temperature values indicate larval concentration were averaged among depths and sites
the arrival of upwelled water in response to increased for each sampling date. Before analysis, the physical
(negative) equatorward wind stress, and increasing temper- variables were normalized (the mean was subtracted from
ature values indicate relaxation conditions. Current veloc- each sample and divided by the standard deviation) and
ity was measured continuously using bottom-mounted larval concentrations were square-root transformed to
acoustic Doppler current profilers in Bodega Bay (RDI deemphasize the contribution of very abundant species.
1200 kHz) and offshore of Bodega Head (RDI 300 kHz): NMDS was conducted on Euclidean distance and Bray–
data were collected in 1-m- and 2-m-depth bins and 3-min Curtis similarity matrices for the physical and biological
and 10-min ensembles, respectively. Velocity ensembles data, respectively. ANOSIM tests (999 maximum permu-
were averaged hourly at both sites and rotated to the tations) determined whether temperature and salinity and
alongshore direction of 332.5u inside the bay and 300.4u the larval assemblages changed with wind condition. The
along the open coast, as defined by the depth-averaged ANOSIM test statistic (R) is a useful measure of separation
principal axis at each site. Each site was characterized by between factors, with zero indicating no separation and one
potential along- and cross-shore transport that was indicating complete separation. A RELATE test was
calculated from time-integrated, depth-averaged water conducted on the physical and biological resemblance
velocities (progressive vector diagrams) over the upper matrices to determine how closely the physical and
30%, middle 30%, and bottom 30% of the water column biological multivariate patterns matched.
(data from the top 10% were omitted due to high errors
associated with side-lobe reflection from the surface). Results
We also mapped surface circulation offshore of the study
site hourly with high-frequency (HF) radar. Pseudo- Water-column structure and current flow during upwelling-
Lagrangian particle trajectories were calculated to deter- relaxation cycles—The eight cruises spanned upwelling-
mine origins of waters in the study region. Water origins relaxation cycles over 7 weeks (Fig. 2A), and the water
were determined following prolonged upwelling and column responded to the fluctuating winds in a similar
relaxation conditions (quasi-steady scenarios). fashion throughout the study area (Fig. 2B,C). Two cruises
Continuous wind data were obtained from a meteoro- (15 Jun, 07 Jul) occurred during upwelling conditions; they
logical buoy, which is located 28 km offshore of Bodega followed prolonged northwesterly winds . 5 m s21
Bay and serves as an index of wind forcing in our region (Fig. 2A), when the water column typically was cold and
(National Data Buoy Center buoy 46013, 38u139300N, well-mixed inside and outside of the bay (Fig. 2B,C). A third
123u199000W; http://las.pfeg.noaa.gov/las6_5/servlets/dataset? upwelling cruise (23 Jun) was conducted following a brief
catitem51708). Upwelling was considered to occur when relaxation period (Fig. 2A), when the water column was still
36-h, low-pass filtered, northwesterly winds were $ 5 m s21 well-mixed and water temperatures were , 10uC, although
and water temperatures were decreasing or , 10uC. winds were briefly from the south at this time (Fig. 2B,C).
Relaxation occurred when northwesterly wind speeds were This cruise was categorized as occurring in upwelling
, 3 m s21, and downwelling occurred when southerly because water-column properties were similar to upwelling
winds exceeded 5 m s21. conditions immediately before the cruise. However, the
surface currents in the bay during this time were not directed
Data analysis—All crustacean larvae and postlarvae equatorward, as they were during the other two upwelling
were identified to species and stage, and counts were cruises (Fig. 2D).
standardized by the volume of water sampled. Larvae of Two more cruises (27 Jun, 29 Jun) followed short periods
most of the 21 taxa collected were combined into five broad of weak, northwesterly winds (Fig. 2A) when the water
taxonomic categories (barnacles, grapsids, pinnotherids, column was beginning to warm and stratify (Fig. 2B,C).
porcellanids, other anomurans) for presentation after Sampling in August occurred following weeks of relaxation
determining that the vertical and horizontal distributions with only brief bouts of weak, northwesterly winds
of species were similar over upwelling-relaxation cycles. (Fig. 2A), and the water column was stratified and warm
Three species are reported individually either because larval (, 13uC; Fig. 2B,C). Three cruises were conducted during
distributions differed between congeners (Cancer antennar- this period: (1) following 2 d of weak, northerly winds (02
ius, C. magister) or only one species was collected within a Aug), (2) following a period of brief poleward winds while
family (Lophopanopeus bellus). Remaining taxa were not the water column was well-stratified (04 Aug), and (3)
abundant and larval distribution patterns were similar to during weak equatorward winds while the water column
the other representative taxa. was more mixed (05 Aug). Warming during relaxation is
We determined how larval distributions for the eight due to both solar heating and advection of warmer offshore
cruises conducted along the three transects inside and three waters into the area (Wing et al. 2003). During brief periods
transects outside the bay differed for the 21 taxa during of northwesterly winds, waters cool due to upwelling and
discrete upwelling and relaxation events using nonmetric mixing, which also reduces stratification.
multidimensional scaling (NMDS) and nonparametric The topographic complexity of the shoreline resulted in
analysis of similarity (ANOSIM). Analyses were conducted different circulation in the bay than on the open coast.460 Morgan et al. Fig. 2. Time-series of physical data sampled in Bodega Bay and off Bodega Head from 07 June to 10 August 2005. (A) Wind (m s21) at National Data Buoy Center 46013—alongshore wind (black) and cross-shore wind (red); gray bars indicate cruise dates; (B) temperature (uC) at 1 m above the bottom (blue) and 1 m below the surface (red) at a mooring on the 10-m isobath inside Bodega Bay; (C) temperature at 1 m (red), 9 m (green), 16 m (blue), 23 m (yellow), and 30 m (black) above the bottom at a mooring on the 30-m isobath off Bodega Head; (D) alongshore current velocity (m s21) at the mooring in Bodega Bay—positive values indicate poleward flow; (E) cross-shore current velocity (m s21) at the mooring in Bodega Bay—positive values indicate onshore flow; (F) alongshore current velocity (m s21) at the mooring off Bodega Head—positive values indicate poleward flow; and (G) cross-shore current velocity (m s21) at the mooring off Bodega Head— positive values indicate onshore flow. Current velocity data were omitted for 10% of the water column near the surface. Near-surface circulation in the bay was typical of wind- previously described by Roughan et al. (2005) and Morgan driven coastal upwelling regions. During upwelling-favorable et al. (2011). conditions, currents were equatorward (Fig. 2D) and In open waters outside the bay, currents were stronger, weakly seaward (Fig. 2E) throughout much of the water but near-surface currents again followed typical patterns of column. When winds relaxed, currents flowed poleward offshore and equatorward flow during upwelling winds (Fig. 2D) and weakly shoreward (Fig. 2E), particularly in (Fig. 2F,G). During relaxation, currents typically were the upper water column. Near-bottom currents in the bay poleward, sometimes extending throughout the water were influenced by subsurface recirculation flow, as column while remaining stronger near the surface, as
Upwelling and larval recruitment 461 Fig. 3. Progressive vector diagrams of currents in (A) in Bodega Bay, and (B) open coastal waters off Bodega Head in 2005. Each line represents the time-integrated velocity from 10 June (5 d before sampling began) to 10 August (5 d after sampling ended). The black line represents velocity averaged over the bottom 30% of the water column, whereas the dark gray and light gray lines represent average velocities for the middle and upper 30% of the water column. Data from the uppermost 10% of the water column were unreliable and omitted. Origins of surface-water parcels calculated from HF-radar data within 5–15 km of Bodega Head on (C) 07 July following a week of upwelling-favorable winds and (D) on 29 June following a few days of weak, relaxation-favorable winds. Crosses represent the origins of surface waters reaching gridded locations along a cross-shelf transect. The gridded locations are reset to a common origin of (0,0) for comparing the direction of flow. The heavy black line represents the orientation of the coast. Due to the limited spatial extent of the HF-radar data, parcels were only ‘back-tracked’ for 2–3 d. previously described by Largier et al. (1993, 2006). Strong upwelling-favorable winds, surface currents outside the bay tidal and diurnal signals were present. Cross-shelf currents typically originated from the northwest (Fig. 3C); during were weaker with a tendency for near-bottom offshore flow relaxation, surface waters originated from the southeast during relaxation periods, as seen previously in this area (Fig. 3D), as previously described (Largier et al. 1993, (Largier et al. 1993; Wing et al. 2003; Dever et al. 2006). 2006; Wing et al. 2003). On average, alongshore transport inside and outside the bay during the summer of 2005 was poleward, with maxima Changes in larval assemblages and distributions during near the bottom in the bay and near the surface in open upwelling-relaxation cycles—Few larvae occurred in rela- waters outside the bay (Fig. 3A,B). Cross-shore transport tively warm, low-salinity water during the eight cruises also differed inside and outside the bay. Inside the bay, that were conducted both inside and outside of the bay, transport alternated between shoreward and seaward near and most of them occurred in cold, saline water during the surface, and it was seaward in the mid- and lower water both relaxation and upwelling conditions (Fig. 4). Very column (Fig. 3A) due to the persistent, deep, recirculation few larvae were collected in waters warmer than 14uC and feature (Roughan et al. 2005; Morgan et al. 2011). Along fresher than a salinity of 33.4, which typically occurred the open coast, net transport was largely poleward and during relaxation. Larval abundance of the eight repre- slightly onshore in the upper water column, with little net sentative taxa in Bodega Bay decreased during two drops transport at the bottom (Fig. 3B). Following prolonged in salinity in June (with the exception of porcellanids on 21
462 Morgan et al.
Fig. 4. Percentage of larvae collected per tow on eight cruises inside and outside Bodega Bay
relative to water temperature and salinity during (A) relaxation and (B) upwelling. Samples were
collected from surface and bottom waters along the six transects. Neuston samples were omitted
because few larvae occurred there during the daytime, and our objective was to compare larval
abundance in different water masses over upwelling-relaxation cycles. Crosses represent tows in
which # 15% of larvae from a survey were collected in one tow.
Jun) and again during a prolonged decrease in salinity on ized by cold, high-salinity water with low chlorophyll
30 July (Fig. 5). fluorescence and turbidity. Little stratification was evident,
The NMDS analysis of temperature and salinity across except in the bay where surface waters were marginally
sites and depths revealed distinct differences among the warmer due to warm water flowing into the bay from
eight cruises that were conducted both inside and outside of Bodega Harbor and Tomales Bay. Larvae of the eight
the bay (Fig. 6A; 2-dimensional stress 5 0). These representative taxa showed a variety of vertical and
differences were separated strongly based on wind forcing horizontal distributions. Larvae of four taxa (barnacles,
(ANOSIM R 5 0.90, p 5 0.018). The two cruises that anomurans, L. bellus, grapsids) occurred deep in the bay,
followed prolonged and strong upwelling winds (15 Jun, 07 and most of them also occurred at a similar depth on the
Jul) and the cruise conducted after a few days of strong open coast (10–12 m). Most pinnotherid and porcellanid
northwesterly winds (23 Jun) were most separated from the larvae occurred outside the bay and were higher in the
dates that were sampled during relaxation conditions. The water column than the other four taxa (5–10 m deep),
two relaxation dates with the least water-column structure although some porcellanid larvae also occurred deep in the
(27 Jun, 05 Aug) were more closely related than the three water column. Cancer antennarius occurred deepest in the
remaining dates that followed prolonged relaxation when water column across the study area, whereas its congener,
the water column was strongly stratified (29 Jun, 02 Aug, C. magister, occurred highest in the water column.
04 Aug). During the onset of relaxation, the water column was
Strong structure in the larval assemblage largely reflected stratified, with relatively warm, low-salinity water occur-
the observed structure in physical conditions (Fig. 6B; ring near the surface (Fig. 8). Chlorophyll fluorescence was
NMDS 2D stress 5 0.05), which was mainly separated by highest in the pycnocline (5–15 m) and the water column
wind forcing (ANOSIM R 5 0.76, p 5 0.018). The larval was less turbid at depth in open coastal waters. Larvae of
assemblages during the three upwelling dates were closely all taxa occurred deep in the water column. Six of the eight
related and well-separated from the five relaxation dates. taxa occurred primarily on the open coast, whereas
Strong groupings occurred within the relaxation dates, pinnotherid and grapsid larvae primarily occurred in the
unlike the little structure observed in the physical NMDS bay.
for those dates. The larval assemblage was similar during 29 Following prolonged relaxation, the water column was
June and 05 August, and it also was alike during 02 August stratified, with relatively warm, low-salinity water, which
and 04 August. The assemblages on these two pairs of dates extended deeper in the water column and farther offshore
differed from that on 27 June. The larval assemblage than during incipient relaxation conditions (Fig. 9). Chlo-
patterns were well-correlated with the temperature and rophyll fluorescence was greatest and turbidity was highest
salinity patterns (RELATE r 5 0.51 p 5 0.014). in the upper water column inside the bay. Larvae of all
The effect of upwelling-relaxation cycles on water- eight taxa were largely absent from the bay and most of
column structure and larval distributions are displayed them occurred 4–5 km from shore. Larvae of four taxa
for three phases of the cycle (Figs. 7–9): upwelling (07 Jul), primarily occurred in the lower water column (barnacles,
early relaxation (29 Jun) and prolonged relaxation follow- pinnotherids, C. antennarius, grapsids), and larvae of three
ing a few days of weakly upwelling-favorable winds (02 taxa occurred higher in the water column (anomurans, L.
Aug). During upwelling, the water column was character- bellus, porcellanids). Only C. antennarius was common inUpwelling and larval recruitment 463 Fig. 5. Mean larval concentration (6 1 SE) of eight representative taxa relative to salinity in Bodega Bay from 07 June to 05 August 2005. surface waters that were warmer than 13uC and a salinity with few larvae occurring in surface waters (anomuran, L. , 33.4. bellus, grapsid, porcellanid, C. antennarius), whereas C. Two days later at night, warm low-salinity water no magister only occurred near the surface. Results were not longer extended to the bottom of the water column due to shown for barnacle and pinnotherid larvae because they do the onset of weak upwelling conditions (Fig. 10A). The not undertake diel vertical migrations (Morgan and Fisher water column was stratified, with warm low-salinity water 2010). Upwelling strengthened by the next night; the water occurring near the surface. Larvae of all eight taxa column was not as strongly stratified and relatively warm, primarily occurred on the inner shelf rather than in the low-salinity water again extended throughout the water bay. Five taxa were concentrated deep in the water column, column (Fig. 10B). Larvae of all six taxa peaked 3 km from
464 Morgan et al.
Fig. 6. Nonmetric multidimensional scaling plots of (A) temperature and salinity averaged
over the three depths and six transects for each of the eight cruises that were conducted during the
daytime inside and outside Bodega Bay, and (B) the concurrent larval assemblages. Dates were
classified as occurring during upwelling or relaxation conditions using wind and temperature data.
shore in the neuston, indicating that they were undertaking and Morgan 2006a; Morgan et al. 2009a). Larvae remained
diel vertical migrations. in cold, saline waters and typically did not enter the warm,
low-salinity waters of the boundary current, being limited
Discussion to bottom waters during early relaxation and offshore
waters during late relaxation events when the boundary
Few larvae occupied the poleward surface current of current spanned the depth of the water column close to
warm, low-salinity water during the three relaxation events, shore. Larval avoidance of the surface lens of warm, low-
indicating that larvae of most species were not transported salinity water was especially apparent when diel vertical
from the retention zone in the lee of Point Reyes to our migrations were suppressed. Only larvae of C. magister
study area. The low concentration of larvae during undertook diel vertical migrations when the boundary
relaxation conditions is consistent with observed patterns current was present, and it is the only crustacean that
of greater recruitment by most species occurring during recruits more during relaxation conditions in some years
prevailing upwelling conditions (Wing et al. 2003; Mace (Wing et al. 2003; Mace and Morgan 2006a). Thus, we haveUpwelling and larval recruitment 465 Fig. 7. Water-column profiles of temperature (uC), salinity, chlorophyll fluorescence, and transmissivity (four upper panels) and larval distributions of eight representative taxa (eight lower panels) across the six transects following prolonged upwelling on 07 July 2005. The water column was profiled at the beginning and end of each transect. Larvae were collected during the daytime at three depths (neuston, surface, bottom) along each transect, as indicated by dots. Neuston samples were omitted because few larvae were collected. Split vertical distributions within taxa largely arose from differences in vertical distributions among species and developmental stages within species. Density structure of the water column was closely aligned with temperature and is not shown. shown how shifting currents over upwelling-relaxation surface waters, where they are advected offshore by Ekman cycles alter larval behavior and displace nearshore larval transport (Morgan et al. 2009c). Postlarvae of these species assemblages, thereby limiting larval recruitment during descend beneath the Ekman layer, where they are advected relaxation events in a region of strong, persistent upwelling. back onshore by upwelling (Morgan et al. 2009c) or rise Intensively profiling larvae in space and time revealed into the neuston, where internal waves or infrequent wind changes in vertical and horizontal distributions over relaxations and reversals may transport them shoreward upwelling-relaxation cycles and complements our previous (Poulin et al. 2002; Morgan et al. 2009c). In turn, the depth research that was focused on prevailing upwelling condi- preferences of postlarvae partially determine the timing of tions. Previously described interspecific differences in the recruitment relative to ocean conditions (Mace and vertical distributions of larvae during prevailing upwelling Morgan 2006a,b; Morgan et al. 2009a). conditions also were evident throughout the study area Larvae of all taxa were much less abundant during during upwelling in the present study (Morgan et al. 2009c; relaxation events, as revealed by the time series of larval Morgan and Fisher 2010). Surface water originated from concentrations in the bay, as well as the plots of larval the northwest during prevailing winds and the vertical concentration on temperature–salinity diagrams (relative to distributions of larvae differed among taxa. These inter- upwelling and relaxation conditions for larvae collected specific differences in vertical distribution determine the both inside and outside the bay). In addition to greatly extent to which larvae complete development nearshore or diminishing larval abundance, relaxation events had a migrate across the continental shelf (Morgan et al. 2009b,c; dramatic effect on larval distributions. In contrast to the Morgan and Fisher 2010). Larvae of many nearshore interspecific differences in vertical distribution during species complete development close to shore by remaining upwelling, widespread interspecific similarity in vertical in waters where cross-shore and alongshore flow is reduced and horizontal distributions of larvae was evident during (Largier et al. 1993; Roughan et al. 2006). They do this by relaxation events. Larvae of all species occurred in the remaining below the shallow Ekman layer at all times or lower water column following the onset of relaxation, with rising into productive surface waters to forage only at night only two taxa occurring in the bay and the other six taxa when offshore flow is weakest (Batchelder et al. 2002; occurring only offshore. Larvae occurred below the lens of Morgan et al. 2009c). Larvae of other species ascend to warm low-salinity water originating from the southeast,
466 Morgan et al. Fig. 8. Water-column profiles of temperature (uC), salinity, chlorophyll fluorescence, and transmissivity (four upper panels) and larval distributions of eight representative taxa (eight lower panels) across the six transects at the onset of relaxation on 29 June 2005. See Fig. 7 for details. Fig. 9. Water-column profiles of temperature (uC), salinity, chlorophyll fluorescence, and transmissivity (four upper panels) and larval distributions of eight representative taxa (seven lower panels) across the six transects following prolonged relaxation on 02 August 2005. Cancer magister larvae were not collected. See Fig. 7 for details.
Upwelling and larval recruitment 467 Fig. 10. Water-column profiles of temperature (uC) and salinity (four upper panels) and larval distributions of eight representative taxa at night (12 lower panels) across the six transects when the water column was (A) strongly stratified (04 Aug 2005) and (B) weakly stratified (05 Aug 2005). Two taxa (barnacles, pinnotherids) were not displayed because they do not undertake diel vertical migrations (Morgan and Fisher 2010). See Fig. 7 for details. which is consistent with the previously described poleward water column was sharply stratified, because larvae did not coastal boundary current (Largier et al. 1993, 2006; Wing undertake diel vertical migrations that previously were et al. 2003). Following a prolonged relaxation event in documented during upwelling conditions (Morgan and spring, warm low-salinity water filled the bay and up to Fisher 2010). Previous field studies also have shown that 3 km from shore. At this time, larvae of all taxa were nearly larval assemblages stay with water masses of particular absent from the bay, and few of them occurred within 3 km temperature and salinity (Banse 1986; Kunze 1995) and from shore. The low-salinity, poleward flow was confined may concentrate near the pycnocline (Metaxas et al. 2009). to a few kilometers from shore, and larvae were more By remaining within their preferred water mass of cold, common and no longer restricted to depth seaward of this saline water during relaxation events, larvae were displaced feature. Interspecific vertical distributions of these larvae seaward in offshore flow even though they could have were similar to those observed nearshore following maintained their vertical distributions by entering the prolonged upwelling events. Thus, the low-salinity near- warm, low-salinity water of the boundary current. Entering surface coastal current filled the bay and inner shelf, the boundary current should not have been overly stressful, displacing larvae offshore into deeper water. because the temperature and salinity change was slight. The absence of larvae in warm, low-salinity surface Remaining with their preferred water mass would keep waters during relaxation events suggests that larvae also larvae of these open-coast species from dispersing toward were absent in the coastal current that originated in the and settling near estuaries, where they would be subject to Gulf of Farallones and not transported poleward from physiologically stressful conditions. Regardless of whether there to our study site. Larvae remained in cold, saline the observed changes in vertical distributions of larvae were water rather than entering this lens of warm, low-salinity behaviorally mediated, larvae of most taxa clearly changed water. This was particularly apparent at night while the in a similar way upon the arrival of the buoyant boundary
468 Morgan et al.
current. In the absence of the distinctive signature of low- and bays (Morgan et al. 2011), as well as straight coastlines
salinity waters, larvae would be prevalent in the surface in upwelling regions. During relaxation events along
layer of the coastal boundary current, making it difficult to upwelling coasts, offshore flow is associated with poleward
detect changes in the vertical and horizontal distributions coastal boundary currents, as has been observed from Point
of larvae in the local assemblage. Arena to Reyes including sites that are well-removed from
Only C. magister larvae occurred in the lens of warm, Bodega Bay (Largier et al. 1993; Wing et al. 2003; Dever
low-salinity water during both nights, presumably because et al. 2006). This offshore flow will transport larvae away
they are more tolerant of these conditions than other from the shoreline, whether it is due to offshore deflection
species. Unlike most of the other study species, C. magister of the poleward flow throughout the water column at
larvae recruit abundantly to the low-salinity waters of San headlands or offshore transport in the bottom Ekman
Francisco Bay (Wild and Tasto 1983). The ability of C. layer. The former process should affect the entire larval
magister to tolerate warm, low-salinity water may explain assemblage at headlands either in the presence or absence
the reason that recruitment of this species is best of a low-salinity surface intrusion, whereas the latter
correlated with the arrival of the coastal boundary current process should have the most pronounced effects on larvae
to our study area (Wing et al. 2003; Mace and Morgan of open-coast species in the presence of a low-salinity
2006a), and it could partially account for the reason that surface intrusion.
C. magister is not well-correlated with the arrival of the In conclusion, a clearly identifiable physical signal
low-salinity, coastal layer every year (Wing et al. 2003; provided insight into the dynamic interaction between
Morgan et al. 2009a). Although larval recruitment is best physics and larval behavior in determining the fate of
correlated with warm, low-salinity water during years larvae in a highly advective upwelling regime. Profiling the
when prolonged relaxation events repeatedly occur (Wing effects of targeted specific oceanographic events on a larval
et al. 2003), it is possible that additional variation could be assemblage can complement shore-based recruitment stud-
accounted for by the amount of precipitation. In years of ies anywhere by providing a mechanistic understanding of
high precipitation, the salinity of the coastal current may observed recruitment patterns (Parnell 2001). Recognizing
be too low for even C. magister larvae to enter, thereby the distinctive signature of different water masses relative
limiting poleward transport to our study site, as may have to current velocities and abundances of species in a larval
occurred during the very wet year of 1995 (Wing et al. assemblage can reveal processes regulating larval recruit-
2003). The timing of the spring transition to peak ment to a study area. Sampling many species concurrently
upwelling conditions also may affect the inter-annual reveals how widespread the effect of the oceanographic
variability in the number of recruiting postlarvae (Shanks event is on the vertical and horizontal distributions of the
and Roegner 2007). The spring transition was late during larval assemblage, and, hence, the ability of multiple species
our study, and C. magister postlarvae did not recruit to regulate their vertical and horizontal positions in
during a companion study that was conducted in our response to shifting ocean conditions. Frequent sampling
study area (Morgan et al. 2009a). After the peak upwelling of larval assemblages complements widespread monitoring
season began, upwelling conditions were typical (Morgan of recruitment from shore and is essential for providing a
et al. 2009a), so that the effect of upwelling-relaxation deeper mechanistic understanding for how shifting ocean
cycles on larval distributions described during our study conditions affect the distribution and abundance of larvae,
would be evident most years. recruitment dynamics and the connectivity of populations
Although larvae remain within preferred water masses along topographically complex coastlines.
and migrate different distances from natal origins by
regulating depth, larval behavior previously was found to Acknowledgments
exert little influence on whether larvae are entrained into We thank S. Miller, L. Katz, M. Sheridan, H. Carson, and T.
retention zones in the lee of headlands (Morgan et al. Clohessey for assisting with field sampling and sample processing.
2011). Larvae of the diverse assemblage of benthic This research was funded by the National Science Foundation
crustaceans accumulated in the lee of Bodega Head during (Biological Oceanography-0326110 and Biological Oceanography-
upwelling conditions when vertically sheared flow arises, 0927196).
regardless of interspecific differences in larval behavior.
Larval behaviors regulating cross-shelf and alongshore References
transport have evolved in response to typical ocean
conditions across species’ ranges and do not appear to BANSE, K. 1986. Vertical distribution and horizontal transport of
facilitate entrainment in the lee of headlands. planktonic larvae of echinoderms and benthic polychaes in an
open coastal sea. Bull. Mar. Sci. 39: 162–175.
The results of our present study are expected to be
BATCHELDER, H. P., C. A. EDWARDS, AND T. M. POWELL. 2002.
repeatable and broadly applicable elsewhere. The effect of Individual-based models of copepod populations in coastal
the boundary current on a diverse larval assemblage was upwelling regions: Implications of physiologically and envi-
observed three times during typical relaxation conditions, ronmentally influenced diel vertical migration on demo-
revealing the process behind the 13-yr record of low graphic success and nearshore retention. Prog. Oceanogr.
recruitment by most species during relaxation events—both 53: 307–333, doi:10.1016/S0079-6611(02)00035-6
inside Bodega Bay and off an exposed headland (Wing BROITMAN, B. R., AND OTHERS. 2008. Spatial and temporal patterns
et al. 2003; Mace and Morgan 2006a; Morgan et al. 2009a). of invertebrate recruitment along the west coast of the United
Our results should apply to the many other small headlands States. Ecol. Monogr. 78: 403–421, doi:10.1890/06-1805.1Upwelling and larval recruitment 469
CALEY, M. J., M. H. CARR, M. A. HIXON, T. P. HUGHES, G. P. ———, ———, AND A. J. MACE. 2009a. Larval recruitment in a
JONES, AND B. A. MENGE. 1996. Recruitment and the local region of strong, persistent upwelling and recruitment
dynamics of open marine populations. Ann. Rev. Ecol. Syst. limitation. Mar. Ecol. Prog. Ser. 394: 79–99, doi:10.3354/
27: 477–500, doi:10.1146/annurev.ecolsys.27.1.477 meps08216
DEVER, E. P., C. E. DORMAN, AND J. L. LARGIER. 2006. Surface ———, ———, ———, L. AKINS, A. M. SLAUGHTER, AND S. M.
boundary layer variability off northern California, USA BOLLENS. 2009b. Cross-shelf distributions and recruitment of
during upwelling. Deep-Sea Res. Part II 53: 2887–2905, crab postlarvae in a region of strong upwelling. Mar. Ecol.
doi:10.1016/j.dsr2.2006.09.001 Prog. Ser. 380: 173–185, doi:10.3354/meps07913
DORMAN, C. E., E. P. DEVER, AND J. L. LARGIER. 2006. Buoy ———, ———, S. H. MILLER, S. T. MCAFEE, AND J. L. LARGIER.
measured wind, wind stress and curl of the wind stress over 2009c. Nearshore larval retention in a region of strong
the shelf off Bodega Bay, California. Deep-Sea Res. Part II upwelling and recruitment limitation. Ecology 90: 3489–3502,
53: 2850–2864, doi:10.1016/j.dsr2.2006.07.006 doi:10.1890/08-1550.1
HUGHES, T. P., A. H. BAIRD, E. A. DINSDALE, N. A. MOLTSCHA- PARNELL, P. E. 2001. The distribution of estuarine and oceanic
NIWSKYJ, M. S. PRATCHETT, J. E. TANNER, AND B. L. WILLIS. water masses on the southern shore of O’ahu, Hawaii:
2000. Supply-side ecology works both ways: The link between Ecological and coastal management implications, and novel
benthic adults, fecundity and larval recruits. Ecology 81: methodology. Limnol. Oceanogr. 46: 1468–1485, doi:10.4319/
2241–2249, doi:10.1890/0012-9658(2000)081[2241:SSEWBW] lo.2001.46.6.1468
2.0.CO;2 POULIN, E., A. T. PALMA, G. LEIVA, D. NARVAEZ, R. PACHECO,
KUNZE, H. B. 1995. Distribution and transport of larvae in the S. A. NAVARETTE, AND J. C. CASTILLA. 2002. Avoiding
lower Hudson River estuary. M.Sc. thesis. State Univ. of New offshore transport of competent larvae during upwelling
York at Stony Brook. events: The case of the gastropod Concholepas concolepas in
LAGOS, N. A., S. A. NAVARRETE, F. VÉLIZ, A. MASUERO, AND J. C. Central Chile. Limnol. Oceanogr. 47: 1248–1255, doi:10.4319/
CASTILLA. 2005. Meso-scale spatial variation in settlement and lo.2002.47.4.1248
recruitment of intertidal barnacles along the coast of central QUEIROGA, H., AND J. BLANTON. 2005. Interactions between
Chile. Mar. Ecol. Prog. Ser. 290: 165–178, doi:10.3354/ behaviour and physical forcing in the control of horizontal
meps290165 transport of decapod Crustacean larvae. Adv. Mar. Biol. 47:
LARGIER, J. L., AND OTHERS. 2006. WEST: A northern California 107–214, doi:10.1016/S0065-2881(04)47002-3
study of the role of wind-driven transport in the productivity ROBERTSON, D. R., D. G. GREEN, AND B. C. VICTOR. 1988. Temporal
of coastal plankton communities. Deep-Sea Res. Part II 53: coupling of production and recruitment of larvae of a Caribbean
2833–2849, doi:10.1016/j.dsr2.2006.08.018 reef fish. Ecology 69: 370–381, doi:10.2307/1940435
———, B. A. MAGNELL, AND C. D. WINANT. 1993. Subtidal ROUGHAN, M., N. GARFIELD, J. LARGIER, E. DEVER, C. DORMAN,
D. PETERSON, AND J. DORMAN. 2006. Transport and retention
circulation over the northern California shelf. J. Geophys.
in an upwelling region: The role of across-shelf structure.
Res. 98: 18147–18179, doi:10.1029/93JC01074
Deep-Sea Res. Part II 53: 2931–2955, doi:10.1016/j.dsr2.
MACE, A. J., AND S. G. MORGAN. 2006a. Biological and physical
2006.07.015
coupling in the lee of a small headland: Contrasting larval
———, A. J. MACE, J. L. LARGIER, S. G. MORGAN, J. L. FISHER,
transport mechanisms in an upwelling region. Mar. Ecol.
AND M. L. CARTER. 2005. Subsurface recirculation and larval
Prog. Ser. 324: 185–196, doi:10.3354/meps324185
retention in the lee of a small headland: A variation on the
———, AND ———. 2006b. Larval accumulation in the lee of a upwelling shadow theme. J. Geophys. Res.-Oceans 110:
small headland: Implications the design of marine reserves. C10027, doi:10.1029/2005JC002898
Mar. Ecol. Prog. Ser. 318: 19–29, doi:10.3354/meps318019 SHANKS, A. L., S. G. MORGAN, J. MACMAHAN, AND A. J. H. M.
METAXAS, A., L. S. MULLINEAUX, AND J. SISSON. 2009. Distribu- REINERS. 2010. Surfzone hydrodynamics as a determinant of
tion of echinoderm larvae relative to the halocline of a salt temporal and spatial variation in larval recruitment. J. Exp.
wedge. Mar. Ecol. Prog. Ser. 377: 157–168, doi:10.3354/ Mar. Biol. Ecol. 392: 140–150, doi:10.1016/j.jembe.2010.
meps07835 04.018
MORGAN, S. G. 1995. Life and death in the plankton: Larval ———, AND C. G. ROEGNER. 2007. Recruitment limitation in
mortality and adaptation, p. 279–321. In L. McEdward [ed.], Dungeness crab populations is driven by variation in
Ecology of marine invertebrate larvae. CRC. atmospheric forcing. Ecology 88: 1726–1737, doi:10.1890/06-
———. 2001. The larval ecology of marine communities, p. 1003.1
159–181. In M. D. Bertness, S. D. Gaines, and M. E. Hay STRATHMANN, R., T. HUGHES, A. KURIS, K. LINDEMAN, S.
[eds.], Marine community ecology. Sinauer. MORGAN, J. PANDOLFI, AND R. WARNER. 2002. Evolution of
———. 2006. Larval migration between the Hudson River estuary local recruitment and its consequences for marine popula-
and New York Bight, p. 157–170. In J. Levinton [ed.], The tions. Bull. Mar. Sci. 70: 377–396.
Hudson River estuary. Oxford Univ. THORSON, G. 1950. Reproductive and larval ecology of marine
———, AND J. R. ANASTASIA. 2008. A behavioral tradeoff bottom invertebrates. Biol. Rev. 25: 1–45, doi:10.1111/j.1469-185X.
conserves transport while increasing the risk of predation 1950.tb00585.x
across the ranges of marine species. Proc. Natl. Acad. Sci. WILD, P., AND R. TASTO. 1983. Life history, environment, and
USA 105: 222–227, doi:10.1073/pnas.0704725105 mariculture studies of the Dungeness crab, Cancer magister,
———, AND J. L. FISHER. 2010. Larval behavior regulates with emphasis on the central California fishery resource. Long
nearshore retention and offshore migration in an upwelling Beach (CA): California Department of Fish and Game Fish
shadow and along the open coast. Mar. Ecol. Prog. Ser. 404: Bulletin, 172.
109–126, doi:10.3354/meps08476 WING, S. R., L. BOTSFORD, L. E. MORGAN, J. M. DIEHL, AND C. J.
———, ———, AND J. L. LARGIER. 2011. Larval entrainment in LUNDQUIST. 2003. Inter-annual variability in larval supply to
the lee of a small headland: Recruitment hotspots along populations of three invertebrate taxa in the northern
windy coasts. Limnol. Oceanogr. 56: 161–178, doi:10.4319/ California Curr. Estuar. Coast. Shelf Sci. 57: 859–872,
lo.2011.56.1.0161 doi:10.1016/S0272-7714(02)00416-X470 Morgan et al. ———, ———, J. S. V. RALSTON, AND J. L. LARGIER. 1998. Associate editor: Christopher M. Finelli Meroplanktonic distribution and circulation in a coastal retention zone of the northern California upwelling system. Limnol. Oceanogr. 43: 1710–1721, doi:10.4319/lo.1998.43.7.1710 YOSHIOKA, P. M. 1982. Role of planktonic and benthic factors in Received: 04 May 2011 the population dynamics of the bryozoan, Membranipora Accepted: 16 December 2011 membranacea. Ecology 63: 457–468, doi:10.2307/1938963 Amended: 10 January 2012
You can also read
